Successful goal-directed behavior requires the ability to select actions that result in reward and avoid actions that result in no reward, or worse, punishment. In the real world, the rules that link actions and their outcomes often change, such that an action that was once rewarded ceases to be so and vice versa. The ability to update decision-making strategy based on positive or negative feedback is the basis for behavioral flexibility. A balance in the weighting of positive and negative feedback's influence over choice selection may be vital for optimal decision-making. It is believed that in disease such as addiction, the ability to learn from negative feedback becomes blunted and reward-seeking overrides normal decision-making processes (Cox et al., 2015; Parvaz et al., 2015). It is well-known that addiction is associated with lower D2 receptor expression in the striatum (Bowirrat et al., 2005; Goldstein and Volkow, 2011; Besson et al., 2013). Previous data from our lab and others implicate D2 receptor-expressing medium spiny neurons (MSNs) of the dorsomedial striatum (DMS) in signaling non-rewarded outcomes and promoting avoidance behavior (Kravitz et al., 2012; Tai et al., 2012). However, the causal role of D2 MSNs in learning from positive and negative outcomes remains unknown. Understanding the contribution of D2 MSN activity to behavioral flexibility will illuminate neural circuit mechanisms that underlie impaired flexibility seen in several psychiatric disorders. I propose to study learning-induced plasticity in the corticostriatal microcircuit to determine if there is a neural signature for behavioral flexibility I will record channelrhodopsin evoked excitatory transmission from the dorsal anterior cingulate (dACC) to the DMS onto D1 or D2 MSNs following the discrimination or reversal phase in a lateralized two-choice decision-making task (Aim 1). Furthermore, I will compare corticostriatal transmission in adults to juveniles, who enact more efficient reversal learning. Next, I will employ chemogenetic tools to selectively inhibit or excite D2 MSNs during a T-maze based spatial reversal task, a 4 choice nonspatial reversal task, and a probabilistic switching task, tha all require flexible updating of decision-making strategies, albeit across different cognitive domains and time-scales (Aim 2). These data will contribute to establishing the striatal circuit mechanisms that support flexible decision-making and reversal learning. By comparing juvenile and adult mice, these experiments will also help to establish how corticostriatal circuits mature during adolescence to alter decision-making strategies. My sponsors and I anticipate that our highly controlled, sensitive, and cell-type specific experimental data from mice will help the larger health community understand the neural basis of impairments in behavioral flexibility. In addition, it may also illuminate a neural basis to developmental changes learning from positive versus negative feedback.